I’ve sat in some interesting Trust and Safety talks that talk about how easy it is for bad actors to set up bot networks to push manipulative agendas. They found recently some of the American major far right content on Reddit doesn’t even come from American accounts

I’m curious to know what tools we have at our disposal to evaluate the level of bot /LLM generated content on platforms we spend our time on.

I’m on Reddit and so many of the first comments are “this is AI slop” that I don’t want to be there anymore

You just know the person who came up with this idea knows they’re an evil prick.

I was looking into tuxedo laptops and system 76, but I have yet to pull the trigger on buying a new laptop. I’m seriously considering it for my next laptop but I’d like something with a solid build quality like my Mac

I take it for granted

You’re not wrong about any of it, it’s just a preconceived notion about Linux I already have that I need to address mentally. Let’s say I was buying a computer, I’d pay for a Microsoft license because I can always fall back on Linux (I don’t even like msft), but Linux was never number one because

To switch out of the idea that putting Linux is something I only do on a shitty old laptop to give it life.

To be honest I could do it now but my computer is not annoying me enough to convince me to do it.

Or maybe the author didn’t break the task down sufficiently before giving it to an LLM?

Neovim lazyvim is very nice imo, it’s great if I’m in a pinch, but I am struggling to make the leap from IntelliJ

When I stare too hard at “looking for the next product” I just find it pigeonholes my creativity a little bit. Atm I’m no such space where I think of solutions. I just want to make stuff that’s different. Right now i picked up Crafting Interpreters

I’ve also heard about nand2tetris and I think I’m going to just venture off into this space for a bit.

I started reading Nystrom’s Crafting Interpreters and it’s just refreshing to write/read code that doesn’t involve plugging a restful endpoint to a controller

Is it really bad that I could be mindlessly delegating to llm and then I have a moment of sobriety where I’m like “wait what??? No! Not like that”

Oh my Goodness, I know exactly that spiral the author has gone down. When you delegate so much to vibe coding your general sense of accomplishment is sapped. You have to balance wanting to make something effective with wanting to learn. It’s crazy

The article mentions a factor could be because agile is in. I think agile makes us more reactive than proactive in bug detection, since we now rely on user reports to detect bugs rather than in house testing proactively discovering them. It’s a fine balance for Saas